180 BRITISH PALAEOZOIC ASTEROZOA. 



tion of the adambulacralia and infero- marginalia from neighbouring arms. It was 

 only lately, when reading some observations of MacBride, that I grasped what this 

 approximation signified. MacBride points out (43, p. 453) that the disc in 

 the great majority of the recent Asteroidea is really composed of adherent arms, 

 and that this is at once made clear when the body-cavity is cut open. " This 

 space is found to be divided up by inwardly projecting folds called interradial 

 septa, which are stiffened by calcareous deposits and represent the conjoined 

 adjacent walls of two arms." He also states (op. cit., p. 475) that in the recent 

 Brisingoidea " the arms have what we must consider to have been the primitive 

 arrangement, since there is no lateral adhesion between them, and interbrachial 

 septa are consequently absent." The Palaeozoic Hudsonasterida? and the Asterozoa 

 of Section C are like the Brisingoidea, and have no lateral adhesion of the arms ; the 

 Mesopalaeasteridae show the beginnings of adhesion by the approximation of the first 

 pair of infero-marginalia which enclose the odontophor, a process which, as we 

 can see by reference to the above-mentioned text-figure, is carried further still in 

 the Promopalseasteridge and the Xenasteridae. In later evolution the odontophor 

 is carried below the surface of the disc and its external origin disguised. In 

 some recent forms, e.g. Brisinga, it still retains its primitive position (compare 

 Perrier, 99, p. 55). The young form of the recent Echinaster sepositus shows an 

 odontophor exactly in the Mesvpalseaster position (see Ludwig 101, PI. X, fig. 8). 

 The infero-marginalia incorporated within the disc by arm-fusion can be traced in 

 the development of Echinaster sepositns (Ludwig, op. cit., PI. X, fig. 11), but they 

 are difficult to recognise in the adult. We can state that as a general rule arm- 

 fusion can be recognised by a study of the external appearance far more readily 

 in the Palaeozoic Asteroidea than in recent forms, where it is only revealed by 

 dissection and exposure of the interradial septa. If we now turn to the recent 

 Ophiuroidea the bases of the arms do not fuse ; in fact they cannot, because they 

 are separated by a secondary growth of the apical parts of the interradii, which 

 grow round on the oral surface to form pouches for the reception of the stomach 

 and for space for the genital bursas (Text-fig. 12, p. 16, and Text-fig. 134, p. 193). 

 Embryological evidence alone shows that this character must have had its 

 inception at a very early stage in the racial history of the Ophiuroidea. We can 

 see this most clearly from the observations on the position of the opening of the 

 water-vascular system to the exterior. Both Ludwig and MacBride have shown 

 that the very young Ophiuroid looks like an Asteroid, and MacBride has also 

 shown that at this stage " the madreporite is on the edge of the disc, and the 

 stone-canal extends horizontally outwards ; and in some Asteroidea there is a 

 similar outward direction in its course" (43, p. 487). The figures and descriptions 

 of Ludwig show the exact point at which the "Ophiuroid impulse" is given. 

 It .is while the madreporite is still in association with the second adambulacral — 

 that is, while the disc is still small (Text-fig. 123). At this very young stage 



